查詢結果分析
相關文獻
- 類神經網路模型應用於超音波換能器與組織參數之估測
- 超音波溫度場之類神經網路模型
- Inverse Estimation of Transducer Parameters with Artificial Neural Network for Ultrasound Hyperthermia
- Estimating One-Dimensional Temperature Field from Insufficient Data During Hyperthermia
- Neural Network Procedures for Taguchi's Dynamic Problems
- A Fast and Efficient Competitive Learning Design Algorithm Based on Weight Vector Training in Transform Domain
- 專家系統振動訊號圖型判別之研究
- 反傳遞模糊類神經網路於流量推估之應用
- 類神經網路(Neural Networks)的種類及其在影像處理上的應用
- C++Fuzzy類神經網路物件導向發展系統之建立
頁籤選單縮合
題名 | 類神經網路模型應用於超音波換能器與組織參數之估測=Estimation of Ultrasound Transducer and Tissue Parameters with Artificial Neural Network |
---|---|
作 者 | 陳永耀; 謝銘鈞; 顏家鈺; 陳啟鴻; 林文灃; | 書刊名 | 中華醫學工程學刊 |
卷期 | 18:2 1998.06[民87.06] |
頁次 | 頁129-137 |
分類號 | 410.1644 |
關鍵詞 | 超音波高溫腫瘤治療; 換能器參數; 血流量; 溫度場; 類神經網路; 反向模型; Ultrasound hyperthermia; Transducer parameters; Perfusion; Temperature distribution; Neural networks; Inverse model; |
語文 | 中文(Chinese) |
中文摘要 | 超音波用來做高溫腫瘤治療時,必需能有效控制其所形成的溫度場,方有可能達 到治療目的而不傷害到正常組織。然而超音波換能器參數與溫度場的函數關係相當複雜,利 用波場壓力方程式和生醫熱傳方程式來計算溫度場是非常耗時的,無論是做療程規劃或做溫 度與系統控制都不是很適合。再者,當確定所要治療的範圍時,必需知道何種超音波換能器 參數可形成所要的溫度場。本研究嘗試利用類神經網路的函數逼近能力來解決這個反向問題 (由溫度場反推換能器參數或血流量 )。 所使用換能器為同心之一外圓環與一內圓盤,外環 內徑為 0.75cm,外徑為 1.5cm, 而內圓盤直徑為 0.75cm,所用之頻率為 1.0MHz,類神經 網路模型中,輸出層為換能器之最大振動速度與相位,而輸入層為加熱組織所形成之溫度場 ;或輸出層為血流量,而輸入層則為換能器參數與溫度場。此類神經網路模型經訓練與測試 後, 結果顯示推測值 (反向模型中換能器參數或血流量所形成的溫度場 ) 與實際計算值相 當接近,且所需時間相當短。 |
英文摘要 | Before we can appropriately employ the ultrasound energy to kill the tumor and not to overheat the normal tissue by using the hyperthermia modality, the temperature distribution generated by the ultrasound heating system must be effectively controlled. However, the relationship between teh transducer parameters and the temperature distribution is rather complicated. Using the acoustic pressure and bioheat transfer equations to calculate teh ultrasound power deposition and the resulting temperature distribution is very time consuming, and hence it is not appropriate for the treatment planning and the system temperature control during the treatment. Furthermore, it is necessary to know the set of ultrasound transducer parameters to form the required temperature distribution when the treatment region is determined. In this paper, artificial neural networks are used to learn the relationship between the transducer parameters (and blood perfusion) and the tissue temperature distribution. The transducer used is a two-element annular array with an inner disk with diameter 0.75 cm, and an outer ring with 0.75 and 1.5 cm for inner and outer diameters, respectively; and the driving frequency is 1.0 MHz. The input layer for the neural network is the tissue temperature distribution, and the output layer is the maximum velocities and the relative phase difference of the transducer elements; or the output layer is blood perfusion, and the input layer is the transducer parameters and the temperature field. Results show that the temperatures predicted by the neural networks are rather close to those calculated by using the acoustic pressure and bioheat transfer equations, and that the computing time for the neural networks is much shorter. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。